BioMed Central
Page 1 of 13
(page number not for citation purposes)
Journal of Translational Medicine
Open Access
Research
A novel role of HLA class I in the pathology of medulloblastoma
Courtney Smith
1,3
, Mariarita Santi
2
, Bhargavi Rajan
1
, Elisabeth J Rushing
4
,
Mi Rim Choi
1
, Brian R Rood
1,3
, Robert Cornelison
5
, Tobey J MacDonald
1,3

and Stanislav Vukmanovic*
1,3
Address:
1
Center for Cancer and Immunology Research, Children's Research Institute, Children's National Medical Center, 111 Michigan Avenue
NW, Washington, DC, USA,

expression, markers of poor prognosis. Peptide- and/or β2m-free forms of MHC class I may
contribute to a more malignant phenotype of medulloblastoma by modulating activation of signaling
molecules such as ERK1/2 that stimulates cell mobility.
Introduction
The host immune system can be harnessed for the treat-
ment of tumors because of the ability of T lymphocytes to
specifically recognize tumor-associated antigens. CD8
+
T
cells destroy tumor cells by perforin-dependent cytotoxic
action, following recognition of MHC class I (HLA in
Published: 12 July 2009
Journal of Translational Medicine 2009, 7:59 doi:10.1186/1479-5876-7-59
Received: 19 March 2009
Accepted: 12 July 2009
This article is available from: http://www.translational-medicine.com/content/7/1/59
© 2009 Smith et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0
),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Journal of Translational Medicine 2009, 7:59 http://www.translational-medicine.com/content/7/1/59
Page 2 of 13
(page number not for citation purposes)
humans) molecules at the cell surface [1]. Structurally, the
MHC class I molecule is comprised of a 44-kD heavy
chain, β2-microglobulin (β2m), and a peptide of 8–10
amino acid residues [2-4]. Presentation of antigenic pep-
tides by MHC class I requires processing involving protea-
some-mediated peptide generation from cytosolic
proteins, peptide transport into the ER mediated by trans-

the outcome of colorectal cancer have been observed in
the same study [18]. The reasons for this counter-intuitive
effect of MHC class I expression remain elusive.
Besides intercellular interactions, MHC class I molecules
have also been implicated in cis modification of signal
transduction. Heavy chains dissociated from β2m and
peptides (called open conformers), and not the fully
assembled MHC class I molecules, associate with a
number of cell surface receptors, resulting in modulation
of their activation [19]. Interestingly, addition of β2m,
which reduces the proportion of open conformers,
appears to either reduce [20] or enhance [21-24] signal
transduction in different experimental systems. Thus, the
direction of signaling modulation by MHC class I open
conformers most likely depends on the identity of the
associated receptor and can be manipulated by exogenous
addition of β2m. Increased serum levels of β2m were
found in several types of cancer [25-32] and were estab-
lished as a poor prognostic factor in bronchial carcinoma
[26], Hodgkin's lymphoma [27], multiple myeloma [31],
renal cell carcinoma [29] and prostate carcinoma [32]. At
present, it is unclear whether the effects of β2m are medi-
ated through HLA class I.
Medulloblastoma is the most common pediatric central
nervous system malignancy accounting for 30% of all
pediatric brain tumors, with the highest prevalence
between the ages of three and eight years [33,34]. Medul-
loblastoma has the tendency to disseminate throughout
the central nervous system early in the course of the dis-
ease. The levels of β2m mRNA are significantly higher in

blocks including: 139 medulloblastomas (47 classic, 40
anaplastic, 25 desmoplastic and 27 with unclassified his-
tology), 21 primitive neuroectodermal tumors, 10 small
cell carcinomas, 5 atypical teratoid rhabdoid tumors, 3
oat cell lung carcinomas, one each of ependymoblastoma
and lymphoma, and 20 tissues of brain metastases from
tumors of various origin. IRB approval was obtained for
the construction and analysis of these tissue microarrays
and all other tumor specimens investigated.
Journal of Translational Medicine 2009, 7:59 http://www.translational-medicine.com/content/7/1/59
Page 3 of 13
(page number not for citation purposes)
Cell Lines
DAOY, D283 and D556 medulloblastoma cell lines were
maintained at 37°C in a humidified atmosphere contain-
ing 5% CO
2
in RPMI media supplemented with 10% Fetal
bovine serum, 2 mM L-glutamine, 1 mM 2-mercaptoetha-
nol, 100 U/ml penicillin and 100 μg/ml streptomycin.
Antibodies
Polyclonal rabbit anti-human β2m and monoclonal
mouse anti-human CD45 (clones 2B11 + PD7/26) were
purchased from DakoCytomation, Carpinteria, CA. The
hybridoma secreting HC-10 antibody, specific for HLA
heavy chain epitope revealed in the absence of β2m and
peptide [38], was kindly provided by Dr. Pan Zheng (Uni-
versity of Michigan School of Medicine, Ann Arbor, MI).
Anti- TAP1 (NOB-1) and TAP2 (NOB-2) antibodies [39]
were provided by Dr. S. Ferrone (Hillman Cancer Center,

nobenzidine (DAB) was the chromogenic substrate. May-
ers hematoxylin was used as a counterstain and followed
with an ammonia wash. Slides were mounted using an
aqueous mounting medium (Faramount, DakoCytoma-
tion). Controls consisted of parallel sections without pri-
mary antibody. Stainings for heavy chain, β2m, TAP1,
TAP2 and CD45 were assigned scores of 0–3 [43] based
on the percentage of stained cells (0- no staining; 1- less
than 10% cells stained; 2- 10–50% cells stained; 3- >50%
cells positive). The staining for c-myc was not restricted to
particular cells, but was rather present or absent through-
out the tumor. Hence, c-myc scores were assigned 0
(absent staining), 1 (weak diffuse staining) or 2 (strong
diffuse staining). Each score was an average of the two
samples graded for each individual tumor in the array.
Each slide was evaluated by three independent graders
including two neuropathologists. The images were
acquired using the following equipment: Microscope –
Carl Zeiss Axioskop; Lenses- Achroplan 40× and Achrop-
lan 20×; Camera – Carl Zeiss Axio Cam HRC; Acquisition
software- Axio Vision Rel. 4.3.
Flow Cytometry
Cells were incubated with either 5% FBS in PBS (controls)
or W6/32 or HC-10 supernatant for 1 hour. After 3 rinses,
cells were incubated with PE labeled donkey anti-mouse
IgG (H+L) (eBiosciences, San Diego, CA) for 30 minutes
at 4°C. Cells were rinsed 3 times and fixed in cytofix (BD
Sciences, San Diego, CA).
RT-PCR
Total cellular RNA, obtained from 10 frozen medulloblas-

Page 4 of 13
(page number not for citation purposes)
Biosolutions, Inc, St. Louis, Missouri) or monoclonal
antibodies for indicated times, at 37°C. Signaling was
halted by the addition of ice cold PBS and cells were lysed
with 1× cell lysis buffer (Cell Signaling, Danvers, Massa-
chussetts) supplemented with PhosSTOP phosphatase
inhibitor cocktail and complete mini protease inhibitor
cocktail (Roche, Indianapolis, Indiana). Lysates were cen-
trifuged at 14,000 rpm for 15 minutes to remove cell
debris, boiled for 5 minutes in loading buffer and sepa-
rated by a 4–12% Bis-Tris Gel (Invitrogen, Carlsbad, Cali-
fornia). Membranes were blocked in 5% milk in TBST for
1 hour at room temperature. Primary and secondary anti-
bodies were diluted in 3% BSA in TBST and incubated
with the membrane at 4°C overnight and at room temper-
ature for 1 hour, respectively. Signal was detected using
the enhanced chemiluminescence system (Pierce, Rock-
ford, IL). Densitometry quantification of the bands was
performed using Quantity One software (Bio-Rad, Her-
cules, CA), according to manufacturer's instructions. The
Band 1/Band 2 ratio was obtained by dividing differences
between intensity units observed in the square areas con-
taining specific bands and an identical blank area drawn
in the immediate vicinity of the band, according to the
following formula: Band 1/Band 2 ratio= (Band 1-blank)/
(Band 2 -blank).
Wound scratch assay
Wound scratch assay was used to evaluate tumor cell
migration [44,45]. DAOY cells were plated in a 60 mm

express MHC class I. The fourth sample (M69) expressed
all four components.
Because of relatively frequent leukocyte infiltration, the
ability to discriminate RNA from tumor versus stromal
host cells was low in many medulloblastoma samples.
IHC analysis of medulloblastoma tissues is advantageous
in this respect, because areas of leukocyte infiltration and
MHC class I expression can be directly visualized and
compared to tumor cell expression in each individual
sample. We therefore analyzed HLA heavy chains, β2m,
TAP1, TAP2 and CD45 intracellular and/or cell surface
expression using medulloblastoma tissue microarrays.
Representative staining patterns for HLA heavy chains are
shown in Fig. 1B. Of the 106 evaluable specimens 87%
showed absent or faint heavy chain positivity (56% scored
0 and 31% scored 1), while scores 2 and 3 were observed
in 5% and 8% of tissues, respectively (Table 1). To gain a
better molecular understanding of MHC class I expres-
sion, medulloblastoma arrays were stained with antibod-
ies to TAP subunits and β2m (Table 1). In agreement with
the mRNA analysis, the majority of medulloblastoma tis-
sues expressed high levels of β2m protein. Surprisingly,
similarly high levels of TAP1 were seen, while staining of
TAP2 was intermediate (Table 1). The 14 samples with
scores 2 or 3 for staining with HC-10 antibody also dis-
played high levels of TAP1, TAP2 and β2m (except for
TAP2 in two samples). Therefore, we conclude that there
is a small subset of medulloblastomas that may express all
components required for functional MHC class I mole-
cules, whereas at least one essential component (heavy

antibody (×40 magnification). Tonsil tissue sections processed identically except for the absence (a) or presence (b) of primary
monoclonal antibody. Examples of medulloblastoma sections graded as 0, 1, 2 and 3.
Journal of Translational Medicine 2009, 7:59 http://www.translational-medicine.com/content/7/1/59
Page 6 of 13
(page number not for citation purposes)
these cases and MHC class I expression appears to be gen-
uinely derived from medulloblastoma cells.
Association of MHC class I expression and anaplastic
medulloblastoma subtype
Anaplastic histopathology [48,49] and c-myc expression
[49-51] are negative prognostic markers for medulloblas-
toma. To test whether MHC class I expression had any
prognostic value for medulloblastomas, we evaluated the
distribution of histological subtypes and expression of c-
myc in specimens with scores 2 or higher for heavy chain
versus those that scored less than 2. The reason for the cut-
off at score 2 is that score 1 was given if only up to 10%
cells in the specimen stained positive. In contrast to the
diffuse character of the anaplasia and c-myc expression,
we considered that the impact, if any, of so few positive
cells on the overall histology of the tumor could not have
been significant.
Of the 106 samples evaluable for HC-10 staining, 31 were
anaplastic, 26 classic, 23 desmoplastic and 26 histologi-
cally unclassified (without evidence of diffuse anaplasia).
Of the fourteen MHC class I-high medulloblastomas 8
were anaplastic, 3 classic, 1 desmoplastic and 2 unclassi-
fied. Similar findings were observed when c-myc expres-
sion was considered. Of the 88 samples that were
evaluable for both HC-10 and c-myc, 59 scored 0, 23

high levels (HC-10
hi
) of HLA class I heavy chains. Fisher exact
test showed statistically significant differences in distribution
with p = 0.0251 for histopathology (n = 105) and p = 0.0257
for c-myc expression (n = 88).
Journal of Translational Medicine 2009, 7:59 http://www.translational-medicine.com/content/7/1/59
Page 7 of 13
(page number not for citation purposes)
thology (n = 105) and 0.0257 for c-myc expression (n =
88).
Binding of exogenous
β
2m alters the balance of open and
closed MHC class I conformers in medulloblastoma cell
line
The levels of open and closed conformer were respectively
analyzed using the HC-10 antibody that detects denatured
heavy chains [38], and W6/32 that binds to the combined
epitope contributed by the α2 and α3 domains of the
heavy chain and β2m, dependent on the presence of pep-
tides in the peptide-binding groove [40-42]. The levels of
open conformers were clearly detectable in DAOY and to
a lesser degree in D556 medulloblastoma cell lines that
display relatively high levels of MHC class I antigens (Fig.
3A). In contrast, open conformers were virtually non-
existent in MHC class I low D283 cell line. To determine
whether β2m can bind to open conformers we evaluated
the effect of HC-10 antibody on β2m binding to DAOY
cells (Fig. 3B). We focused on DAOY cells because they

DAOY cells reduced the relative levels of open, and
increased the levels of closed conformers (Fig. 3B), sug-
gesting that binding of exogenous β2m can alter the bal-
ance between the open and closed conformers.
Engagement of open MHC class I conformers modulates
phosphorylation of ERK1/2
We next examined whether altering balance between the
open and closed MHC class I conformers may contribute
to signal modulation. We therefore evaluated the impact
of exogenous β2m on phosphorylation of ERK1/2 and
AKT that are downstream arms of many receptor path-
ways. Increased levels of phospho-ERK1/2 were found
15–30 minutes following addition of exogenous β2m to
DAOY cells (Fig. 4A), while the levels of phospho-AKT
remained largely unaltered. The optimal increase in
ERK1/2 phosphorylation was achieved with 2 μg/ml β2m
(Fig. 4B), which is well within the range of physiological
concentration in human serum [26,27,29]. Further, the
effect was partially inhibited by anti-β2m antibodies
despite their 10-fold molar deficit (Fig. 4C). Because there
is a constitutive low level of ERK1/2 phosphorylation in
the absence of any treatment, we refer to the effect of β2m
as modulation, rather than induction of ERK1/2 phos-
phorylation. Consistent with the available levels of open
conformers, increased levels of pERK1/2 were seen in
D556 cells, albeit with a slightly delayed kinetics (Fig.
4D), but not in MHC class I deficient D283 cells (Fig. 4E).
We next tested whether HC-10 antibody could prevent
β2m -induced ERK1/2 phosphorylation. We were sur-
prised to see that adding HC-10 itself modified ERK1/2

ml), and 1.44- (0.08 μg/ml) fold increase in pERK1/2 over the
background observed in the absence of β2m. C) β2m (2 μg/
ml) was mixed with anti-β2m antibody (0.2 mg/ml) prior to
the addition to DAOY cells and ERK1/2 activation was ana-
lyzed 15 minutes post β2m addition. Densitometric quantifi-
cation is shown below indicates 43% inhibition by anti-β2m
antibodies of β2m-induced ERK1/2 phosphorylation. D-E)
The effect of β2m (2 μg/ml) incubation of indicated time
lengths on phosphorylation of ERK1/2 in D556 (D) or D283
(E) cells was examined by Western blotting. F) Summary of
pERK1/2 quantification in three different cell lines at indi-
cated times after β2m treatment, relative to the levels in
untreated cells (results are expressed as fold induction).
Journal of Translational Medicine 2009, 7:59 http://www.translational-medicine.com/content/7/1/59
Page 9 of 13
(page number not for citation purposes)
showed marginal ability to invade, the presence of β2m
clearly induced significant migratory activity (Fig. 6A).
Evaluation of the surface area of the scratch at 0 and 24
hours time points indicated that untreated cells recolo-
nized only 6%, while the β2m- and HC-10-treated recov-
ered 24% and 22% of the wound area, respectively (Fig.
6B). The migration was inhibited in the presence of 100
μm PD98059, pharmacologic inhibitor of upstream
member of the ERK1/2 activation pathway [52]. Thus, the
ERK1/2 activation enhanced by the engagement of open
HLA class I conformers may contribute to higher migra-
tory capacity of medulloblastoma cells.
Discussion
We show herein that classical MHC class I is undetectable

negative prognostic impact of a loss of MHC class I in
small cell lung carcinoma, pancreatic carcinoma, cervical
cancer, colon cancer and melanoma is consistent with this
view [9-12]. Alternatively, HLA-negative may have arisen
by transformation of cells that originally did not express
MHC class I. These examples would include well-differen-
tiated cells, such as muscle cells and neurons, unless the
expression is induced by inflammatory stimuli. Because
medulloblastoma develops from neuronal precursors and
the CNS is relatively protected from the immune system,
we consider selection by CD8
+
T cells an unlikely cause of
the HLA-negative phenotype of most medulloblastomas.
An explanation for HLA class I expression by a minority of
medulloblastomas remains to be established.
The expression of c-myc and/or presence of anaplasia,
which are negative prognostic markers for medulloblast-
oma [48-51], were associated with HLA class I expression.
Thus, HLA class I expression may be associated with more
aggressive medulloblastomas. If the association is con-
firmed in larger studies, MHC class I expression may prove
to be an important biomarker of the malignant pheno-
type. Because follow-up clinical data were not available
for all patients in this study, we were unable to make cor-
relation between high HLA class I expression and disease
outcomes. Nevertheless, the association of higher expres-
sion of β2m with metastatic disease in a previous study of
ours [35] is consistent with the findings of the present
study. Despite considerable evidence suggesting that the

ecule on its own, the literature is replete on its involve-
ment in signaling. Thus, cross-linking MHC class I
molecules by antibodies in Jurkat T cells and T cell clones
induces TCR activation similar to that induced by TCR
engagement [53,54]. Neither cytoplasmic nor transmem-
brane domain of heavy chains are required for signal
transmission, suggesting that MHC class I molecules asso-
ciate and use the signal transduction machinery of other
cell surface receptors [55]. Further, addition of β2m can
reduce [20] or enhance [21-24] signal transduction by dif-
ferent receptors (Fig. 7), confirming the ability of open
MHC class I conformers of modifying the function of var-
ious cell surface receptors [19]. ERK1/2 activation in
medulloblastomas can occur following activation of
growth factor receptors, such as EGFR, PDGFR, IGF1R and
CXCR4 [56-59]. However, we found no increase in phos-
phorylation of these receptors following engagement of
the open conformers (data not shown). Consistent with
this notion is also the fact that we observed no consistent
increase in phosphorylation of Akt that is normally acti-
vated by EGFR, PDGFR, IGF1R and CXCR4 receptors.
Alternatively, the asymmetric signaling may result from
the selective (in)action of protein phosphatase 2A
(PP2A). PP2A is activated in medulloblastomas [60] and
can selectively inactivate Akt or ERK1/2 due to binding of
distinct regulatory subunits [61]. Thus, the exact mecha-
nism of HLA class I open conformer engagement-medi-
ated activation of ERK1/2 in medulloblastoma remains to
be determined.
Dysregulation of ERK1/2 has been implicated in tumori-

on individual cell surface receptor signaling and/or by the
presence or absence of an overriding tumor-specific CD8+
T cell response.
Abbreviations
β2m: β2-microglobulin; IHC: immunohistochemistry;
TAP: transporter associated with antigen processing.
Competing interests
The authors declare that they have no competing interests.
Schematic representation of signal transduction modification by MHC class I open conformers in medulloblastomaFigure 7
Schematic representation of signal transduction modification by MHC class I open conformers in medulloblas-
toma. The closed conformation of MHC class I (left) is composed of the heavy chain possessing three extracellular domains
(α1, α2 and α3) non-covalently bound to β
2
m (β2) and antigen peptide (pep). Dissociation of β
2
m and peptide leads to the for-
mation of open conformation that can interact with receptors (R) on the cell surface. This interaction dulls the impact of
receptor ligand (L) binding (here symbolically represented by binding of one unit of ligand per one unit of receptor; note, how-
ever, that the impact of open conformers may not be related to ligand binding, but some other mechanism, i.e. preventing opti-
mal receptor conformation). Binding of extracellular β
2
m to open conformer releases the receptor, enabling full blown
signaling indicated by increased phosphorylation (P) of intracellular portion of the receptor.
Journal of Translational Medicine 2009, 7:59 http://www.translational-medicine.com/content/7/1/59
Page 12 of 13
(page number not for citation purposes)
Authors' contributions
CS carried out IHC staining (except for c-myc), immun-
ofluorescence staining, PCR, Western blot and migration
experiments, analyzed results and participated in the writ-

class I and class II molecules. Curr Opin Immunol 1995, 7:85-96.
3. Pamer E, Cresswell P: Mechanisms of MHC class I–restricted
antigen processing. Annu Rev Immunol 1998, 16:323-358.
4. Vukmanovic S, Lilic M, Santori FR, Demaria S, Kulig K: Peptide load-
ing of nascent MHC class I molecules. Arch Immunol Ther Exp
(Warsz). 2001, 49(3):195-201.
5. York IA, Rock KL: Antigen processing and presentation by the
class I major histocompatibility complex. Annu Rev Immunol
1996, 14:369-396.
6. Koopmann J-O, Hammerling GJ, Momburg F: Generation, intracel-
lular transport and loading of peptides associated with MHC
class I molecules. Curr Opin Immunol 1997, 9:80-88.
7. Cabrera T, Lopez-Nevot MA, Gaforio JJ, Ruiz-Cabello F, Garrido F:
Analysis of HLA expression in human tumor tissues. Cancer
Immunol Immunother 2003, 52:1-9.
8. Chang CC, Ferrone S: Immune selective pressure and HLA
class I antigen defects in malignant lesions. Cancer Immunol
Immunother 2007, 56:227-236.
9. Delp K, Momburg F, Hilmes C, Huber C, Seliger B: Functional defi-
ciencies of components of the MHC class I antigen pathway
in human tumors of epithelial origin. Bone Marrow Transplant
2000, 25:S88-95.
10. Agrawal S, Kishore MC: MHC class I gene expression and regu-
lation. J Hematother Stem Cell Res 2000, 9:795-812.
11. Nacht M, Dracheva T, Gao Y, Fujii T, Chen Y, Player A, Akmaev V,
Cook B, Dufault M, Zhang M, et al.: Molecular characteristics of
non-small cell lung cancer. Proc Nat Acad Sci USA 2001,
98:15203-15208.
12. Kamarashev J, Ferrone S, Seifert B, Böni R, Nestle FO, Burg G, Dum-
mer R: TAP1 down-regulation in primary melanoma lesions:

receptors and its effect on the insulin-signaling cascade. Mol
Biol Cell 1997, 8:2463-2474.
21. Rowley DR, Dang TD, McBride L, Gerdes MJ, Lu B, Larsen M: Beta-
2 microglobulin is mitogenic to PC-3 prostatic carcinoma
cells and antagonistic to transforming growth factor beta 1
action. Cancer Res 1995, 55:781-786.
22. Paczek L, Czarkowska-Paczek B, Korczak-Kowalska G, Wierzbicki P,
Bartlomiejczyk I, Górski A: Involvement of beta2-microglobulin
in CD69 expression on T cells. Arch Immunol Ther Exp (Warsz).
2001, 49(3):239-242.
23. Nomura T, Huang W, Zhau H, Wu D, Xie Z, Mimata H, Zayzafoon
M, Young AN, Marshall FF, Weitzmann MN, et al.: Beta2-
microglobulin promotes the growth of human renal cell car-
cinoma through the activation of the protein kinase A, cyclic
AMP-responsive element-binding protein, and vascular
endothelial growth factor axis. Clin Cancer Res 2006,
12:7294-7305.
24. Nomura T, Huang WC, Seo S, Zhau H, Mimata H, Chung LW: Tar-
geting beta2-microglobulin mediated signaling as a novel
therapeutic approach for human renal cell carcinoma. J Urol
2007, 178:292-300.
25. Bunning RA, Haworth SL, Cooper EH: Serum beta-2-microglob-
ulin levels in urological cancer. J Urol 1979, 121:624-625.
26. Hallgren R, Nou E, Lundqvist G: Serum beta 2-microglobulin in
patients with bronchial carcinoma and controls. Cancer 1980,
45:780-5.
27. Dimopoulos MA, Cabanillas F, Lee JJ, Swan F, Fuller L, Allen PK, Hage-
meister FB: Prognostic role of serum beta 2-microglobulin in
Hodgkin's disease. J Clin Oncol 1993, 11:1108-11.
28. Sadamori N, Mine M, Hakariya S, Ichiba M, Kawachi T, Itoyama T,

(page number not for citation purposes)
33. Roberts RO, Lynch CF, Jones MP, Hart MN: Medulloblastoma: a
population-based study of 532 cases. Journal of Neuropathology
and Experimental Neurology 1991, 50(2):134-44.
34. Packer RJ: Childhood medulloblastoma: progress and future
challenges. Brain Dev 1999, 21:75-81.
35. MacDonald TJ, Brown KM, LaFleur B, Peterson K, Lawlor C, Chen Y,
Packer RJ, Cogen PH, Stephan DA: Expression profiling of medul-
loblastoma: PDGFRA and the RAS/MAPK pathway as thera-
peutic targets for metastatic disease. Nat Genet 2001,
29:143-152.
36. Abouantoun TJ, Macdonald TJ: Imatinib blocks migration and
invasion of medulloblastoma cells by concurrently inhibiting
activation of platelet-derived growth factor receptor and
transactivation of epidermal growth factor receptor. Mol
Cancer Ther 2009 in press.
37. Kononen J, Bubendorf L, Kallioniemi A, Bärlund M, Schraml P,
Leighton S, Torhorst J, Mihatsch MJ, Sauter G, Kallioniemi OP: Tissue
microarrays for high-throughput molecular profiling of
tumor specimens. Nat Med 1998, 4:844-847.
38. Perosa F, Luccarelli G, Prete M, Favoino E, Ferrone S, Dammacco F:
Beta 2-microglobulin-free HLA class I heavy chain epitope
mimicry by monoclonal antibody HC-10-specific peptide. J
Immunol 2003, 171:1918-1926.
39. Albers A, Abe K, Hunt J, Wang J, Lopez-Albaitero A, Schaefer C,
Gooding W, Whiteside TL, Ferrone S, DeLeo A, et al.: Antitumor
activity of human papillomavirus type 16 E7-specific T cells
against virally infected squamous cell carcinoma of the head
and neck. Cancer Res 2005, 65:11146-11455.
40. Maziarz RT, Fraser J, Strominger JL, Burakoff SJ: The human HLA-

gen-processing machinery in medulloblastoma. Cancer Res
2007, 67:5471-5478.
48. Perry A: Medulloblastomas with favorable versus unfavorable
histology: how many small blue cell tumor types are there in
the brain? Adv Anat Pathol 2002, 9:345-350.
49. Ellison D: Classifying the medulloblastoma: insights from
morphology and molecular genetics.
Neuropathol Appl Neurobiol
2002, 28:257-282.
50. Stearns D, Chaudhry A, Abel TW, Burger PC, Dang CV, Eberhart CG:
c-Myc overexpression causes anaplasia in medulloblastoma.
Cancer Res 2006, 66:673-681.
51. Gulino A, Arcella A, Giangaspero F: Pathological and molecular
heterogeneity of medulloblastoma. Curr Opin Oncol 2008,
20:668-675.
52. Dudley DT, Pang L, Decker SJ, Bridges AJ, Saltiel AR: A synthetic
inhibitor of the mitogen-activated protein kinase cascade.
Proc Natl Acad Sci USA 1995, 92:7686-7689.
53. Geppert TD, Wacholtz MC, Patel SS, Lightfoot E, Lipsky PE: Activa-
tion of human T cell clones and Jurkat cells by cross-linking
class I MHC molecules. J Immunol 1989, 142:3763-3772.
54. Skov S, Bregenholt S, Claesson MH: MHC class I ligation of
human T cells activates the ZAP70 and p56lck tyrosine
kinases, leads to an alternative phenotype of the TCR/CD3
zeta-chain, and induces apoptosis. J Immunol 1997,
158:3189-3196.
55. Gur H, el-Zaatari F, Geppert TD, Wacholtz MC, Taurog JD, Lipsky
PE: Analysis of T cell signaling by class I MHC molecules: the
cytoplasmic domain is not required for signal transduction. J
Exp Med 1990, 172:1267-1270.